Explicit pattern matching assignment algorithm

摘要: Sharing data between two tracking systems frequently involves use of an object map: the transmitting system sends a frame with multiple observations, and receiving uses assignment algorithm to correlate information its local observation base. The usual prescription for this problem is optimal (such as JVC or auction) using cost matrix based upon chi-squared distances remote data. does not actually perform pattern matching, so approach robust large registration errors when there exist differences in number observations held by both systems. Performance new that function including terms track random presented: explicitly includes bias sets thus provides maximum likelihood solution problem. In practice, near perfect accuracy cases where exceed dimension transmitted map mismatches numbers made This performance extends many methodology produces nearly 100% time. paper theoretical foundation solved comparison achieved existing approaches.